I would love to see a study on the history of wellbeing and suffering. This is perhaps more challenging as our understanding of how people suffered in the past is (arguably) poorly understood (as is our understanding of exactly how/why people suffer today!). But a first order approximation could look at generic factors that we expect to correlate with wellbeing, such as the proportion of people living in slavery or servitude; the personal freedoms people had; the levels of violence; and so on. Then a more detailed study—which would probably require expertise beyond history—would be to look at more direct (but harder to find historically) indicators of wellbeing such as mental health, self-reported happiness, suicide rates, etc.
Yeah, I share the view that either more research on that topic or a summary of existing for EAs would be valuable. (I imagine a lot of relevant work on that already exists, but I’ve been wrong about such things before, and in any case it could be good for someone to read it and extra the most EA-relevant insights.)
I think I’d see this as an (important) subset of “1. The history of various types of growth and progress (economic, intellectual, technological, moral, political, etc.)”. Would you agree?
(That wouldn’t negate the value of your comment—many of the topics I listed are very broad, and this post becomes more useful to people if commenters break them down into more specific topics, suggest ways they could be investigated, etc.)
I agree that the “first approximation” I mentioned—looking at generic factors that we expect to correlate with wellbeing, such as slavery or servitude, personal freedoms, violence—would be a subset of “1. The history of various types of growth and progress...”.
But I feel like a more detailed investigation of wellbeing/suffering through history lies outside of “1. The history of various types of growth and progress...”. I say this because what we call “progress” does not necessarily correlate with wellbeing/suffering. And I think this *might* lead charities and movements such as EA to potentially overestimate the effects of intuitively useful interventions. I should add that this is potentially speculative and controversial! But I feel that there are important questions that haven’t been fully tackled: Does growth really improve wellbeing? Does increasing life expectancy really reduce suffering, or does it make people overly sensitive to death? Were people in previous centuries—where violence and disease were high—as unhappy as we’d expect if we just look at these factors? Or are there more subtle factors that affect happiness?
Sometimes I feel like “progress” is about “satisfying people’s stated preferences” rather than “making people happy”. And what we think we want isn’t always what makes us happy!
So rather than looking directly at violence, growth, death rates, etc, (which I expect has been done many times), I’d like to see a detailed study that looks at more direct indicators of wellbeing such as mental health, self-reported happiness, suicide rates—and many more. And then a comparison between this and the usual “progress” studies. Perhaps this has also been done though and I’ve missed it.
Anyway I’d be very interested to hear what you think as I’ve not properly discussed these ideas before!
I have a draft, which I’ll hopefully publish in the coming weeks/months, on “Will humanity achieve its full potential, as long as existential catastrophe is prevented?”
I think an argument in favour of “Yes” is that it might be highly likely that, if we don’t suffer an existential catastrophe, there will be positive trends across the long-term future in all key domains. And I think that that argument could in turn be supported by the argument that such trends have been the norm historically, or that human agency will ensure such positive trends.
So I thought a bit about how true that seems to be. I’ll quote the relevant part of the draft, as it seems somewhat relevant here. (Note that I’m not an expert, and barely even did any googling; this was based on intuitions and what I happened to already know/believe.)
---
I believe there’s strong evidence that there have been positive trends in many domains in many periods and places before the Industrial Revolution. Relevant domains may include violence levels, the size of people’s moral circles, and use of reason and scientific thinking.
See e.g. The Better Angels of Our Nature.
I believe there’s some evidence that this represents a fairly widespread pattern. But I’m less certain of that. And there’ve definitely been “negative” trends in certain domains, times, and places (e.g., [insert example here; I have some ideas but should Google them]).
I believe there’s strong evidence that, since sometime around the Industrial Revolution, there have been positive trends across most of the world and in most domains that matter.
But even since the Industrial Revolution, there have been at least some negative trends or stagnation in some domains, times, and places. And these might include someof the “most important” domains, times, and places in relation to evaluating the FINE hypothesis.
Here are some plausibly important domains where I think there’s at least some evidence of negative trends recently in the developed world:
Human-caused animal suffering (especially on factory farms)
Political discourse
Political polarisation
Respect for science, scientists, and/or truth
Mental health [maybe also suicide rates? should google this]
Drug abuse
Incarceration rates (perhaps especially or only in the US)
Economic inequality
There were also some negative trends in particular domains, times, and places that were later reversed, but seem like they plausibly could’ve become quite lastingly bad. E.g., various trends in Germany and Russia leading up to and during WWII.
And there are plausibly important domains for which I’m not aware of evidence of substantial progress recently (e.g., democratisation in China).
Overall, I think historical trends are more consistent than inconsistent with the [argument that, if we don’t suffer an existential catastrophe, there will be positive trends across the long-term future in all key domains]. But that the matter isn’t totally clear-cut, and would likely benefit from much more detailed analysis.
“Will humanity achieve its full potential, as long as existential catastrophe is prevented?”
I think an argument in favour of “Yes” is that it might be highly likely that, if we don’t suffer an existential catastrophe, there will be positive trends across the long-term future in all key domains.
That there will be positive trends doesn’t necessarily entail that humanity (or some other entities) will achieve its full potential, however. It’s possible that the future will be better than the present, without humanity achieving its full potential. And the value difference between such a future and a future where humanity achieves its full potential may be vast.
I agree that there is an historical argument for positive future trends, but it seems that one needs additional steps to conclude that humanity will achieve its full potential.
Yeah, I definitely agree. This was part of my motivation for writing that draft. (Also, even if just “positive trends” was enough—which I agree that it isn’t—finding that werepositive trends in the past doesn’t guarantee there willbepositive trends in the future.)
More broadly, my impression is that some EAs are very confident the answer to the titular question is “Yes”, and I feel like I haven’t seen very strongarguments for such high confidence.
The draft is not necessarily arguing in favour of “Yes” (or “No”) overall; it’s primarily intended to highlight the question and stimulate and scaffold discussion.
(Happy to share the draft, if you or others are interested.)
Stepping back first: I’m quite morally uncertain, but the moral theory I have the highest degree of belief in is “something like classical hedonistic utilitarianism, with a moral circle that includes basically sentient beings, across any point in time”. (My moral circle therefore may or may not include mammals, insects, digital minds, etc., depending on whether they “empirically turn out to be sentient”—though it’s quite unclear what that means. For expected value reasons, concerns about digital minds, insects, etc. play a substantial role in my priorities.)
The classical hedonistic utilitarianism bit (setting aside the moral circles bit) makes me very strongly inclined to agree that:
what really matters is how (human) wellbeing has changed over time, and
that it’s unfortunate that discussion/studies of “growth” and “progress” often focus on things that may not be strongly correlated with (human) wellbeing.
I’d say the focus is, as you suggest, often on “satisfying people’s stated preferences”. But I’d even go further and say that it’s often on one of the following things:
what the person in this discussion or doing this study thinks is a typical or ideal preference
what that person themselves thinks is terminally valuable (regardless of preferences)
whatever is easiest to measure/discuss and seems plausibly related somehow to wellbeing, preferences, or valuable things
Two EAs who’ve done what seems to me good work in relation to subjective wellbeing, its measurement, and its correlation with other things are Michael Plant and Derek Foster. (Though I don’t think they focused much on history.)
...but then there’s the moral circles bit. This makes me think that (a) human wellbeing is unlikely to be a dominating concern, and (b) wellbeing at the moment or so far is unlikely to be a dominating concern.
So I care about present-day human wellbeing primarily to the extent that it correlates with across-all-time, across-all-sentient-life wellbeing. And this means that, for instrumental reasons, I probably actually should pay more attention to other proxies, like GDP or technological developments, than to wellbeing. (This doesn’t mean it’s clear to me that GDP growth or technological developments tend to be good, but that they’re likely important, for good or ill. See differential progress.)
So, in contrast to what I might have said a few years ago when my moral circle hadn’t expanded to consider nonhumans and future beings more, I wouldn’t personally be extremely excited about historical analysis of changes in human wellbeing over time, and what affected those changes. But:
I think that’d be quite exciting from a human-centric, non-longtermist perspective
I think it’s still net-positive, and maybe quite positive, from my perspective, because understanding this may help us make various predictions about important aspects of the future, and work out how we should intervene
I’ll sort-of elaborate on this in a separate comment
You’ve made some really good points here and I agree with most of it! And we’re on the same page in terms of “hedonistic utilitarianism, with a moral circle that includes basically all sentient beings, across any point in time”.
I guess my main motivation for wanting to see a historical study of well-being is because I feel that, to fully understand what makes humans happy, it is valuable to consider a wide range of possible human life experiences. Studying history does this: we can consider a wide range of societies, lifestyles, circumstances etc, and ask which humans were happy and which were suffering. And comparing this to standard “progress” measures such as violence and life expectancy can help us understand whether interventions to improve such measures are the best we can do. Then this can help us design and implement future strategies to improve well-being moving forward.
Great post!
I would love to see a study on the history of wellbeing and suffering. This is perhaps more challenging as our understanding of how people suffered in the past is (arguably) poorly understood (as is our understanding of exactly how/why people suffer today!). But a first order approximation could look at generic factors that we expect to correlate with wellbeing, such as the proportion of people living in slavery or servitude; the personal freedoms people had; the levels of violence; and so on. Then a more detailed study—which would probably require expertise beyond history—would be to look at more direct (but harder to find historically) indicators of wellbeing such as mental health, self-reported happiness, suicide rates, etc.
Thanks!
Yeah, I share the view that either more research on that topic or a summary of existing for EAs would be valuable. (I imagine a lot of relevant work on that already exists, but I’ve been wrong about such things before, and in any case it could be good for someone to read it and extra the most EA-relevant insights.)
I think I’d see this as an (important) subset of “1. The history of various types of growth and progress (economic, intellectual, technological, moral, political, etc.)”. Would you agree?
(That wouldn’t negate the value of your comment—many of the topics I listed are very broad, and this post becomes more useful to people if commenters break them down into more specific topics, suggest ways they could be investigated, etc.)
I agree that the “first approximation” I mentioned—looking at generic factors that we expect to correlate with wellbeing, such as slavery or servitude, personal freedoms, violence—would be a subset of “1. The history of various types of growth and progress...”.
But I feel like a more detailed investigation of wellbeing/suffering through history lies outside of “1. The history of various types of growth and progress...”. I say this because what we call “progress” does not necessarily correlate with wellbeing/suffering. And I think this *might* lead charities and movements such as EA to potentially overestimate the effects of intuitively useful interventions. I should add that this is potentially speculative and controversial! But I feel that there are important questions that haven’t been fully tackled: Does growth really improve wellbeing? Does increasing life expectancy really reduce suffering, or does it make people overly sensitive to death? Were people in previous centuries—where violence and disease were high—as unhappy as we’d expect if we just look at these factors? Or are there more subtle factors that affect happiness?
Sometimes I feel like “progress” is about “satisfying people’s stated preferences” rather than “making people happy”. And what we think we want isn’t always what makes us happy!
So rather than looking directly at violence, growth, death rates, etc, (which I expect has been done many times), I’d like to see a detailed study that looks at more direct indicators of wellbeing such as mental health, self-reported happiness, suicide rates—and many more. And then a comparison between this and the usual “progress” studies. Perhaps this has also been done though and I’ve missed it.
Anyway I’d be very interested to hear what you think as I’ve not properly discussed these ideas before!
I have a draft, which I’ll hopefully publish in the coming weeks/months, on “Will humanity achieve its full potential, as long as existential catastrophe is prevented?”
I think an argument in favour of “Yes” is that it might be highly likely that, if we don’t suffer an existential catastrophe, there will be positive trends across the long-term future in all key domains. And I think that that argument could in turn be supported by the argument that such trends have been the norm historically, or that human agency will ensure such positive trends.
So I thought a bit about how true that seems to be. I’ll quote the relevant part of the draft, as it seems somewhat relevant here. (Note that I’m not an expert, and barely even did any googling; this was based on intuitions and what I happened to already know/believe.)
---
I believe there’s strong evidence that there have been positive trends in many domains in many periods and places before the Industrial Revolution. Relevant domains may include violence levels, the size of people’s moral circles, and use of reason and scientific thinking.
See e.g. The Better Angels of Our Nature.
I believe there’s some evidence that this represents a fairly widespread pattern. But I’m less certain of that. And there’ve definitely been “negative” trends in certain domains, times, and places (e.g., [insert example here; I have some ideas but should Google them]).
I believe there’s strong evidence that, since sometime around the Industrial Revolution, there have been positive trends across most of the world and in most domains that matter.
See e.g. The world is much better; The world is awful; The world can be much better, Three wild speculations from amateur quantitative macrohistory, and Enlightenment Now.
But even since the Industrial Revolution, there have been at least some negative trends or stagnation in some domains, times, and places. And these might include some of the “most important” domains, times, and places in relation to evaluating the FINE hypothesis.
Here are some plausibly important domains where I think there’s at least some evidence of negative trends recently in the developed world:
Human-caused animal suffering (especially on factory farms)
Political discourse
Political polarisation
Respect for science, scientists, and/or truth
Mental health [maybe also suicide rates? should google this]
Drug abuse
Incarceration rates (perhaps especially or only in the US)
Economic inequality
There were also some negative trends in particular domains, times, and places that were later reversed, but seem like they plausibly could’ve become quite lastingly bad. E.g., various trends in Germany and Russia leading up to and during WWII.
And there are plausibly important domains for which I’m not aware of evidence of substantial progress recently (e.g., democratisation in China).
Overall, I think historical trends are more consistent than inconsistent with the [argument that, if we don’t suffer an existential catastrophe, there will be positive trends across the long-term future in all key domains]. But that the matter isn’t totally clear-cut, and would likely benefit from much more detailed analysis.
That there will be positive trends doesn’t necessarily entail that humanity (or some other entities) will achieve its full potential, however. It’s possible that the future will be better than the present, without humanity achieving its full potential. And the value difference between such a future and a future where humanity achieves its full potential may be vast.
I agree that there is an historical argument for positive future trends, but it seems that one needs additional steps to conclude that humanity will achieve its full potential.
Yeah, I definitely agree. This was part of my motivation for writing that draft. (Also, even if just “positive trends” was enough—which I agree that it isn’t—finding that were positive trends in the past doesn’t guarantee there will be positive trends in the future.)
More broadly, my impression is that some EAs are very confident the answer to the titular question is “Yes”, and I feel like I haven’t seen very strong arguments for such high confidence.
The draft is not necessarily arguing in favour of “Yes” (or “No”) overall; it’s primarily intended to highlight the question and stimulate and scaffold discussion.
(Happy to share the draft, if you or others are interested.)
Thanks, yes I’d be interested.
Ok, I’ve sent you a message :)
I think these are good points.
Stepping back first: I’m quite morally uncertain, but the moral theory I have the highest degree of belief in is “something like classical hedonistic utilitarianism, with a moral circle that includes basically sentient beings, across any point in time”. (My moral circle therefore may or may not include mammals, insects, digital minds, etc., depending on whether they “empirically turn out to be sentient”—though it’s quite unclear what that means. For expected value reasons, concerns about digital minds, insects, etc. play a substantial role in my priorities.)
The classical hedonistic utilitarianism bit (setting aside the moral circles bit) makes me very strongly inclined to agree that:
what really matters is how (human) wellbeing has changed over time, and
that it’s unfortunate that discussion/studies of “growth” and “progress” often focus on things that may not be strongly correlated with (human) wellbeing.
I’d say the focus is, as you suggest, often on “satisfying people’s stated preferences”. But I’d even go further and say that it’s often on one of the following things:
what the person in this discussion or doing this study thinks is a typical or ideal preference
what that person themselves thinks is terminally valuable (regardless of preferences)
whatever is easiest to measure/discuss and seems plausibly related somehow to wellbeing, preferences, or valuable things
Two EAs who’ve done what seems to me good work in relation to subjective wellbeing, its measurement, and its correlation with other things are Michael Plant and Derek Foster. (Though I don’t think they focused much on history.)
...but then there’s the moral circles bit. This makes me think that (a) human wellbeing is unlikely to be a dominating concern, and (b) wellbeing at the moment or so far is unlikely to be a dominating concern.
So I care about present-day human wellbeing primarily to the extent that it correlates with across-all-time, across-all-sentient-life wellbeing. And this means that, for instrumental reasons, I probably actually should pay more attention to other proxies, like GDP or technological developments, than to wellbeing. (This doesn’t mean it’s clear to me that GDP growth or technological developments tend to be good, but that they’re likely important, for good or ill. See differential progress.)
So, in contrast to what I might have said a few years ago when my moral circle hadn’t expanded to consider nonhumans and future beings more, I wouldn’t personally be extremely excited about historical analysis of changes in human wellbeing over time, and what affected those changes. But:
I think that’d be quite exciting from a human-centric, non-longtermist perspective
I think it’s still net-positive, and maybe quite positive, from my perspective, because understanding this may help us make various predictions about important aspects of the future, and work out how we should intervene
I’ll sort-of elaborate on this in a separate comment
You’ve made some really good points here and I agree with most of it! And we’re on the same page in terms of “hedonistic utilitarianism, with a moral circle that includes basically all sentient beings, across any point in time”.
I guess my main motivation for wanting to see a historical study of well-being is because I feel that, to fully understand what makes humans happy, it is valuable to consider a wide range of possible human life experiences. Studying history does this: we can consider a wide range of societies, lifestyles, circumstances etc, and ask which humans were happy and which were suffering. And comparing this to standard “progress” measures such as violence and life expectancy can help us understand whether interventions to improve such measures are the best we can do. Then this can help us design and implement future strategies to improve well-being moving forward.